Release history and notes
*************************

Sequence based identifiers are used for versioning (schema follows
below):

   major.minor[.revision]

* It's always safe to upgrade within the same minor version (for
  example, from 0.3 to 0.3.4).

* Minor version changes might be backwards incompatible. Read the
  release notes carefully before upgrading (for example, when
  upgrading from 0.3.4 to 0.4).

* All backwards incompatible changes are mentioned in this document.


0.13.2
======

2026-03-06

* Widen protol ranges and improve protocol matching in general.

* Clean up.

* Minor documentation fixes.

* Tested against Python 3.14 and 3.15.


0.13.1
======

2025-05-20

* Tested against Python 3.13.

* Tested against Python 3.12.

* Add documentation tests.

* Updated bundled tld names.

* Updated some tests to reflect changes in tld names.

* Fixes in docs.

* Merge *pytest*, *mypy* and *coverage* configs into *pyproject.toml*.

* Drop Python 3.7 and 3.8 support. Minimum required version now is
  Python 3.9.


0.13
====

2023-02-28

* Drop Python 2.7, 3.5 and 3.6 support. Minimum required version now
  is Python 3.7.


0.12.7
======

2023-02-01

* Make sure to fail silently on bad URL patterns.

* Tested against Python 3.11.

* Tested against Python 3.10.

* Updated bundled tld names.


0.12.6
======

2021-06-05

* Move "Registry" class from "tld.registry" to "tld.base".

* Reformat code using "black".

* Log information on updated resources of the "update_tld_names".


0.12.5
======

2021-01-11

Note:

  Release dedicated to defenders of Armenia and Artsakh (Nagorno
  Karabakh) and all the victims of Turkish and Azerbaijani aggression.

* Fixed lower-cased *parsed_url* attributes (*SplitResult*) when
  getting tld as object (*as_object=True*).


0.12.4
======

2021-01-02

* Tested against Python 3.9.


0.12.3
======

2020-11-26

* Separate parsers for (a) public and private and (b) public only
  domains. This fixes a bug. If you want an old behaviour:

  The following code would raise exception in past.

     from tld import get_tld

     get_tld(
         'http://silly.cc.ua',
         search_private=False
     )

  Now it would return *ua*.

     get_tld(
         'http://silly.cc.ua',
         search_private=False
     )

  If you want old behavior, do as follows:

     from tld.utils import MozillaTLDSourceParser

     get_tld(
         'http://silly.cc.ua',
         search_private=False,
         parser_class=MozillaTLDSourceParser
     )

  Same goes for "get_fld", "process_url", "parse_tld" and "is_tld"
  functions.


0.12.2
======

2020-05-20

* Add mozilla license to dist.

* Fix MyPy issues.


0.12.1
======

2020-04-25

Note:

  In commemoration of Armenian Genocide.

* Correctly handling domain names ending with dot(s).


0.12
====

2020-04-19

* Use Public Suffix list instead of deprecated Mozilla's MXR.


0.11.11
=======

2020-03-10

* Minor speed-ups, reduce memory usage.


0.11.10
=======

2020-02-05

* Python 2.7 and 3.5 fixes.


0.11.9
======

2019-12-16

* Adding test TLDs list to the package.


0.11.8
======

2019-12-13

* Minor fixes in setup.py.


0.11.7
======

2019-12-13

Note:

  There have been no code changes since 0.11.2. The only change is
  that support for Python 2.7 and 3.5 has been added.

* Added support for Python 2.7.


0.11.6
======

2019-12-12

* Targeted releases for all supported Python versions.


0.11.5
======

2019-12-12

* Targeted releases for all supported Python versions.


0.11.4
======

2019-12-12

* Changed order of the releases (Python 3.6 and up come first, then
  Python 3.5).

* Make all distributions except Python 3.5 universal.


0.11.3
======

2019-12-12

* Added missing resources to the Python 3.5 release.


0.11.2
======

2019-12-12

* Bring back Python 3.5 support.


0.11.1
======

2019-12-11

* Minor speed ups.

* More on adding typing.


0.11
====

2019-12-09

Note:

  Since introduction of parser classes, usage of "NAMES_SOURCE_URL"
  and "NAMES_LOCAL_PATH" of the "tld.conf" module is deprecated. Also,
  "tld_names_local_path" and "tld_names_source_url" arguments are
  deprecated as well. If you want to customise things, implement your
  own parser (inherit from "BaseTLDSourceParser").

* Drop support for Python versions prior to 3.6.

* Clean-up dependencies.

* Introduce parsers.

* Drop "tld_names_source_url" and "tld_names_local_path" introduced in
  the previous release.

* Minor speed-ups (including tests).


0.10
====

2019-11-27

Note:

  This is the last release to support Python 2.

* Make it possible to provide a custom path to the TLD names file.

* Make it possible to free up some resources occupied due to loading
  custom tld names by calling the "reset_tld_names" function with
  "tld_names_local_path" parameter.


0.9.8
=====

2019-11-15

* Fix for occasional issue when some domains are not correctly
  recognised.


0.9.7
=====

2019-10-30

Note:

  This release is dedicated to my newborn daughter. Happy birthday, my
  dear Ani.

* Handling urls that are only a TLD.

* Accepts already splitted URLs.

* Tested against Python 3.8.


0.9.6
=====

2019-09-12

* Fix for update-tld-names returns a non-zero exit code on success
  (introduced with optimisations in 0.9.4).

* Minor tests improvements.


0.9.5
=====

2019-09-11

* Tests improvements.


0.9.4
=====

2019-09-11

* Optimisations in setup.py, tests and console scripts.

* Skip testing the update-tld-names functionality if no internet is
  available.


0.9.3
=====

2019-04-05

* Added *is_tld* function.

* Docs updated.

* Upgrade test suite.


0.9.2
=====

2019-01-10

* Fix an issue causing certain punycode TLDs to be deemed invalid.

* Tested against Python 3.7.

* Added tests for commands.

* Dropped Python 2.6 support.

* TLD source updated to the latest version.


0.9.1
=====

2018-07-09

* Correctly handling nested TLDs.


0.9
===

2018-06-14

Note:

  This release contains backward incompatible changes. You should
  update your code.The "active_only" option has been removed from
  "get_tld", "get_fld" and "parse_url" functions. Update your code
  accordingly.

* Removed "active_only" option from "get_tld", "get_fld" and
  "parse_url" functions.

* Correctly handling exceptions (!) in the original TLD list.

* Fixes in documentation.

* Added "parse_tld" function.

* Fixes the "python setup.py test" command.


0.8
===

2018-06-13

Note:

  This release contains backward incompatible changes. You should
  update your code.Old "get_tld" functionality is moved to "get_fld"
  (first-level domain definition). The "as_object" argument (False by
  default) has been deprecated for "get_fld".

     res = get_tld("http://www.google.co.uk", as_object=True)

  **Old behaviour**

     In: res.domain
     Out: 'google'

     In: res.extension
     Out: 'co.uk'

     In: res.subdomain
     Out: 'www'

     In: res.suffix
     Out: 'co.uk'

     In: res.tld
     Out: 'google.co.uk'

  **New behaviour**

     In: res.fld
     Out: 'google.co.uk'

     In: res.tld
     Out: 'co.uk'

     In: res.domain
     Out: 'google'

     In: res.subdomain
     Out: 'www'

  When used without "as_object" it returns "co.uk".**Recap**If you
  have been happily using old version of "get_tld" function without
  "as_object" argument set to "True", you might want to replace
  "get_tld" import with "get_fld" import:

     # Old
     from tld import get_tld
     get_tld('http://google.co.uk')

     # New
     from tld import get_fld
     get_fld('http://google.co.uk')

* Move to a Trie to match TLDs. This brings a speed up of 15-20%.

* It's now possible to search in public, private or all suffixes (old
  behaviour). Use "search_public" and "search_private" arguments
  accordingly. By default (to support old behavior), both are set to
  "True".

* Correct TLD definitions.

* Domains like ******.xn--fiqs8s* are now recognized as well.

* Due to usage of "urlsplit" instead of "urlparse", the initial list
  of TLDs is assembled quicker (a speed-up of 15-20%).

* Docs/ directory is included in source distribution tarball.

* More tests.


0.7.10
======

2018-04-07

* The "fix_protocol" argument respects protocol relative URLs.

* Change year in the license.

* Improved docstrings.

* TLD source updated to the latest version.


0.7.9
=====

2017-05-02

* Added base path override for local .dat file.

* *python setup.py test* can used to execute the tests


0.7.8
=====

2017-02-19

* Fix relative import in non-package for update-tls-names script. #15

* "get_tld" got a new argument "fix_protocol", which fixes the missing
  protocol, having prepended "https" if missing or incorrect.


0.7.7
=====

2017-02-09

* Tested against Python 3.5, 3.6 and PyPy.

* pep8 fixes.

* removed deprecated *tld.update* module. Use "update-tld-names"
  command instead.


0.7.6
=====

2016-01-23

* Minor fixes.


0.7.5
=====

2015-11-22

* Minor fixes.

* Updated tld names file to the latest version.


0.7.4
=====

2015-09-24

* Exposed TLD initialization as "get_tld_names".


0.7.3
=====

2015-07-18

* Support for wheel packages.

* Fixed failure on some unicode domains.

* TLD source updated to the latest version.

* Documentation updated.


0.7.2
=====

2014-09-28

* Minor fixes.


0.7.1
=====

2014-09-23

* Force lower case of the URL for correct search.


0.7
===

2014-08-14

* Making it possible to obtain object instead of just extracting the
  TLD by setting the "as_object" argument of "get_tld" function to
  True.


0.6.4
=====

2014-05-21

* Softened dependencies and lowered the "six" package version
  requirement to 1.4.0.

* Documentation improvements.


0.6.3
=====

2013-12-05

* Speed up search


0.6.2
=====

2013-12-03

* Fix for URLs with a port not handled correctly.

* Adding licenses.


0.6.1
=====

2013-09-15

* Minor fixes.

* Credits added.


0.6
===

2013-09-12

* Fixes for Python 3 (Windows encoding).


0.5
===

2013-09-13

* Python 3 support added.


0.4
===

2013-08-03

* Tiny code improvements.

* Tests added.